Patients with KPD represent a unique subset of type 2 diabetes in which obese/overweight patients present with diabetic ketoacidosis despite absence of humoral or cellular ? cell autoimmunity, implying unique and profound ? cell dysfunction of unknown. In this project we propose to test the general hypothesis that the severe but partially reversible and intermittent ? cell dysfunction in KPD is mediated through diminished availability of arginine for NO synthesis, which in turn impairs insulin secretion. Specific hypotheses to be tested are: 1. Compared to healthy subjects and usual, non-ketosis-prone T2D patients, KPD patients have decreased intracellular arginine availability due to a) increased arginine catabolism caused by elevated plasma arginase activity, and b) decreased arginine synthesis due to diminished citrulline synthesis and efflux from the intestine. 2. Due to diminished arginine availability, KPD patients have impaired glucose-stimulated insulin secretion, as well as impaired endothelial function; however the insulin response to arginine + glucose is normal. 3. Enhancement of intracellular arginine availability will result in improved glucose-stimulated insulin secretion in KPD patients. However, due to increased arginine catabolism by arginase, this can be achieved more effectively by dietary citrulline supplementation than by dietary arginine supplementation. Using stable isotope tracer methods these hypotheses will be tested in groups of KPD patients, type 2 diabetics and healthy controls by achieving the following aims: Aim 1a. Measure under basal conditions and following a hyperglycemic clamp, a) arginine flux, de novo synthesis from citrulline, release from protein breakdown, rate of conversion to ornithine and to NO; b) ornithine and citrulline fluxes (to calculate arginine hydrolysis and syntheses of arginine and NO; c) phenylalanine flux as a measure of protein breakdown. Aim 1b. Measure, under conditions of fasting and feeding: the rate of conversion of glutamine to citrulline in the intestine, the rate of splanchnic uptake of arginine, its oxidation and its conversion to ornithine as an index of arginase activity in the intestine, and in vitro plasma arginase activity. Aim 2. Measure glucose- and arginine-stimulated insulin secretion and arterial flow- mediated dilation, before and following a hyperglycemic clamp. Aim 3. In three groups of KPD patients, measure arginine flux, synthesis from citrulline, NO synthesis and the measurements described in Aim 2, before and after supplementation with citrulline, arginine or isonitrogenous alanine (placebo). Data generated from these studies could have high impact for understanding the cause of ? cell dysfunction not only in KPD but also in all forms of type 2 diabetes. Further, if citrulline supplementation improves glucose- stimulated insulin secretion and endothelial function to a greater extent than arginine, it will provide a saf and inexpensive adjunct therapy for KPD and type 2 diabetes.